Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rubbish "Rubbish"

Author: Chessfun

Date: 19:33:28 04/29/02

Go up one level in this thread


On April 29, 2002 at 21:25:21, Ratko V Tomic wrote:

>Note that a single game result (3 possible values)
>contains 1.58 bits of information. The game
>Thorsten posted had 239 plies. If you take
>(the commonly used estimate) an average of
>35 legal moves at each ply, the 239 plies
>then contain 239*log2(35)=1226 bits of
>information.
>
>Thus, by retaining only the score, you would need to
>play 773 games to accumulate as much information
>as available in the full 239 ply sequence.
>
>Additionally, a human analysis by a knowledgable
>player is not only more interesting to read than
>391:382, but the conceptual building blocks of the
>human analysis are seeds/patterns for creative thought,
>vastly more useful for other humans, be it players
>looking for strategy to use against the programs
>or programmers looking for ways to improve their
>programs.


Indeed human analysis may be good if done by a knowledgeable person. Now lets
examine Thorsten's.

Header. it seems fritz7 gets out-evaluated :-))

Message. this is a typical example of a game where fritz says 0.00 all the time.

If that's what you consider  "a human analysis by a knowledgable player is not
only more interesting to read than 391:382".........so be it. Personally I'll
stick with numbers.

Sarah.




This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.